SKAN for beginners and other dynamics

You may have heard quite a bit about SKAN lately. But what exactly is it? SKAN is the framework developed by Apple for attributing installations in mobile applications. The main goal is to help measure campaign activity without revealing the identity of users. In the case of SKAN, the App Store itself acts as an intermediary in the measurement, and the attribution process is carried out using Apple’s technology.

In short, when a user downloads from an Ad, Apple provides basic information to the advertiser to indicate if the installation resulted from the Ad without revealing who made the download, thus protecting user privacy.

The most well-known version is SKAN 3, which continues to be the most used. However, the industry has gradually transitioned to SKAN 4, which offers more measurement and optimization data.

Unlike version 3.0, which only received one postback, with SKAN 4.0, we have up to 3 postbacks, each corresponding to a specific attribution window.

The so-called lockWindow:

• Postback 1: Attribution 0 to 2 days. Delay of 24 to 48 hours.
• Postback 2: Attribution 3 to 7 days. Delay of 24 to 144 hours.
• Postback 3: Attribution 8 to 35 days. Delay of 24 to 144 hours.

The main advantage of lockWindow is the reduced waiting time for a postback. However, shortening its duration also reduces the information available in each measurement window. The lock does not apply to inactive users, which could distort the data sets toward more active or higher-value users. 

(Source: Apple Developer)

With each new version, SKAN offers much more value but becomes more difficult and challenging. There are many factors involved, but having an overall view of what to define and how to act can help us familiarize ourselves with a change that will be inevitable.

Tips to boost your strategy

1. Verticals: SKAN 4.0 allows measuring events for up to 35 days with three different postbacks after the app’s initial installation. So, if you work with subscription or streaming verticals, you can use it.
   
   This is especially useful for applications with longer funnels, as it allows the measurement of user quality metrics, such as conversions after a free trial with several weeks’ retention.
2. Reports: It is important to thoroughly understand how reports work, considering that the source ID can have 2, 3, or 4 digits and can be returned in different dimensions in each postback.
   Advertisers, developers, and media agencies must understand the new players well to act accordingly and not affect the data, such as LTV reports, which tend to be more sensitive.
3. Optimize: To carry out successful campaigns, it is crucial to make decisions in advance. With SKAN 4.0, unless the lockWindow function is used, the first postback will arrive 3 or 4 days after installation.
   
   Focus on what you really need to measure, considering that there is a time between obtaining user activity and updating the data. The installation can be recorded several days later, which can affect the strategies you have planned.

For more information on What’s new with SKAdNetwork, click here.